This invention relates to methods of preparing shape selective zeolite catalysts, such as ZSM-5, and methods for their use to synthesize hydrocarbons such as olefins, in particular by conversion of lower monohydric alcohols and/or their ether derivatives.
Olefins, especially ethylene and propylene, are used on a large scale as intermediates for the manufacture of staple products such as olefin polymers, ethylene oxide, non-ionic detergents, glycols and fibre-forming polyesters. Processes for producing olefins usually involve non-catalytic pyrolysis of volatile hydrocarbons such as natural gas liquids or petroleum distillates. Catalytic pyrolysis processes have been proposed but do not appear to have reached industrial use.
In countries where such volatile hydrocarbons are not accessible but such feedstocks as coal, oil shale and methane, and consequently carbon monoxide/hydrogen synthesis gas derived therefrom, are available, it would be desirable to produce olefins from synthesis gas. It has been proposed to do this by converting the synthesis gas to methanol or to hydrocarbons and/or their oxygenated derivatives and reacting such products over shape selective acidic zeolites, e.g., of the ZSM-5 family. (See for example U.S. Pat. Nos. 3,894,106; 3,894,107; 4,025,571; and 4,052,479).
Shape selective zeolite materials, both natural and synthetic, have been demonstrated in the past to have catalytic capabilities for various types of organic compound conversions. These materials are ordered porous crystalline metalosilicates (e.g. aluminosilicates) having a definite crystalline structure within which there are a large number of cavities and channels, which are precisely uniform in size. Since the dimensions of these pores are such as to accept, for adsorption, molecules of certain dimensions while rejecting those of larger dimensions, these materials are deemed to possess the property of shape selectivity, have been referred to as "molecular sieves", and are utilized in a variety of ways to take advantage of these properties.
Such shape selective zeolites include a wide variety of positive ion-containing crystalline aluminosilicates, both natural and synthetic. Aluminosilicates can be described as a rigid three-dimensional network of SiO.sub.4 and AlO.sub.4 in which the tetrahedra are cross-linked by the sharing of oxygen atoms whereby the ratio of the total valence of the tetrahedra-containing aluminum is typically balanced by the inclusion in the crystal of a cation, for example an alkali metal or an alkaline earth metal cation. This can be expressed by formula wherein the ration of Al to the number of various cations, such as Ca/2, Sr/2, Na, K, or Li is equal to unity. One type of cation may be exchanged either in entirety or partially by another type of cation utilizing ion exchange techniques in a conventional manner. By means of such cation exchange, it has been possible to vary the size of the pores in a given aluminosilicate by suitable selection of the particular cation. The spaces between the tetradehra are occupied by molecules of water prior to dehydration.
A preferred group of shape selective crystalline aluminosilicates, designated as those of the ZSM-5 type (e.g. see U.S. Pat. No. 3,702,886) is well known for use in the synthesis of olefins from syn gas derived materials such as methanol. Other shape selective zeolite materials are also well known for this purpose as discussed in the aforedescribed patents.
Unfortunately, the use of shape selective zeolites to catalyze methanol and/or dimethyl ether conversion for olefin production is not entirely satisfactory because such zeolites are also well known to catalyze the formation higher hydrocarbons from the initially produced olefins such as C.sub.5 +paraffins, naphthenes, aromatics and alkylated aromatics. The particular distribution of products obtained from the use of any given catalyst is typically controlled by the reaction conditions, particularly temperature. Thus, while there is not a clear line of demarcation in product distribution as a function of temperature, it has been recognized (for example see U.S. Pat. No. 3,894,107) that as the reaction temperature is increased, the methanonl conversion can be shifted in favor of the formation of ethers, olefins, aromatics and alkylated aromatics at respectively higher reaction temperatures. The use of temperature control to influence product distribution is illustrated in U.S. Pat. Nos. 4,052,429 and 4,058,576 wherein staging of the reactions is employed. The partial pressure of the reactant feed has also been observed to influence olefin selectivity. Thus, U.S. Pat. No. 4,025,576 discloses the use of a subatmospheric partial pressure of the reactant feed to improve its conversion with enhanced olefin selectivity. Subatmospheric partial pressure of the reactant feed is obtained either by maintaining a partial vacuum in the conversion zone, or by co-feeding a diluent. Suitable diluents include any substantially inert substance that is a gas or vapor at reaction temperature such as steam, as well as nitrogen, carbon dioxide, carbon monoxide, hydrogen, and the like.
While optimization of operating conditions for a given zeolite to optimize a desired product distribution is important, such procedures are limited in the effects which can be produced thereby by inherent limitations in the physical and chemical properties of the zeolite.
Zeolite catalytic properties can be strongly influenced by such factors as crystal morphology, uniformity of crystal morphology, acidity characteristics and silica/alumina mole ratio, cation identity, pore size distribution, degree of crystallinity, as well as by control of numerous process conditions employed during the preparation of the zeolite which in turn can affect one or more of the aforedescribed characteristics in addition to producing indeterminate effects. Thus, the number of permutations and combinations of possible preparative process conditions, and resulting catalyst characteristics, is astronomical. Consequently, one is faced with a seal of variables in attempting to correlate a particular set of catalyst properties, a means for consistently achieving these properties, and the ultimate effect of a given set of properties on catalyst performance.
Furthermore, it will be understood that catalyst performance includes not only catalyst activity, and selectivity to a particular product distribution but also catalyst life.
For example, olefin synthesis reactions inevitably are accompanied by complex side reactions such as aromatization, polymerization, alkylation and the like to varying degrees. As a result of these complex reactions, a carbonaceous deposit is laid down on the catalyst which is referred to by petroleum engineers as "coke". The deposit of coke on the catalyst tends to seriously impair the catalyst efficiency for the principal reaction desired, and to substantially decrease the rate of conversion and/or the selectivity of the process. Thus, it is common to remove the catalyst from the reaction zone after coke has been deposited thereon and to regenerate it by burning the coke in a stream of oxidizing gas. The regenerated catalyst is returned to the conversion stage of the process cycle. The period of use between catalyst regenerations is often referred to as catalyst life. In short, coke deposits are believed to be a primary contributing factor to reductions in catalyst life. There are obvious economic incentives to improve the catalyst life such as the savings in capital investment for regeneration equipment.
As with such catalyst properties as activity and selectivity, one can control catalyst life through control of the operating conditions. However, it would be a significant advantage if catalyst life could be improved by improving the nature of the catalyst itself through its preparative procedure.
Unfortunately, it is very difficult to predict improvements in catalyst performance from variations in conventional methods of synthesis. This stems from the fact that the most conventional way to identify a particular zeolite is by its characteristic X-ray diffraction pattern. However, catalyst performance of two zeolites with the same XRD pattern can differ drastically, in many instances for indeterminate reasons. One is therefore forced to search beyond the XRD pattern of a zeolite capable of enhancing catalyst performance.
The present invention focuses on the combination of active pH control with an acid to adjust the initial pH of the reaction mixture within specifically defined limits, in the presence of sodium cations and at a specifically defined SiO.sub.2 /Al.sub.2 O.sub.3 mole ratio conducive to the synthesis of olefins from methanol.
In the article "Crystallization of Zeolite ZSM-5 From a Single Cation System", by H. Nakamoto, and H. Takahashi, Chemical Letters, pp. 1739-1742 (1981), the authors report the production of a ZSM-5 zeolite in the presence of TPA.sup.+ as the only cation, by control of the concentration ratios of (TPA).sub.2 O/SiO.sub.2, SiO.sub.2 /Al.sub.2 O.sub.3, and H.sub.2 O/SiO.sub.2 in the reaction mixture. They conclude that the crystallization rate is strongly dependent on the (TPA).sub.2 O/SiO.sub.2 mole ratio and indicate that a minimum ratio of 0.2 is necessary for the formation of ZSM-5 in their system (e.g. SiO.sub.2 /Al.sub.2 O.sub.3 =100; Temp.=150.degree. C.; H.sub.2 O/SiO.sub.2 =81; Na/SiO.sub.2 =0.0038), thereby ensuring sufficient alkalinity in the reaction mixture to induce dissolution of the amorphous solid to form soluble active species from which nuclei grow. Since crystallization was observed to occur rapidly after the induction period for dissolution, the formation of nuclei is suggested as the rate determining, step in the overall process. No crystalline phase is observed after 5 days at a (TPA).sub.2 O/SiO.sub.2 ratio of 0.1. Increasing the SiO.sub.2 /Al.sub.2 O.sub.3 ratio increased crystallization independent of whether single TPA.sup.+ or binary Na.sup.+ /TPA.sup.+ system is employed. However at SiO.sub.2 /Al.sub.2 O.sub.3 ratios below 100, the Na.sup.+ /TPA.sup.+ system achieves better crystallization than the TPA.sup.+ -system, while at ratios above 100 the Na cation is said not to play an important role in crystallization. Increasing the H.sub.2 O/SiO.sub.2 ratio was found to decrease the crystallization rate. Finally, as the (TPA).sub.2 O/SiO.sub.2 and SiO.sub.2 /Al.sub.2 O.sub.3 ratios were increased in a mono cation TPA.sup.+ system, larger well defined crystals were observed to form having a barrel shape. Catalyst performance is not reported for any of the synthesized zeolites.
It is appropriate to mention that Nakamoto et al as well as many of the hereinafter discussed articles mention the "alkalinity" of the reaction mixture. The concept of increasing or decreasing alkalinity is to be distinguished from increasing or decreasing pH. When relatively strong bases such as NaOH or TPAOH are present in the initial reaction mixture, the pH of the same will almost always be 14. Thus, as more base is added while the alkalinity may increase, the initial pH will remain at 14.
For example, while FIG. 1 of Nakamoto et al illustrates increasing alkalinity, the initial reaction mixture pH of all of the runs is 14.
A paper by K. Chao, T. Tasi, and M. Chen, entitled "Kinetic Studies on the Formation of Zeolite ZSM-5", Journal of Chem. Soc. Trans. 1, Vol. 77 pp. 547-55 (1981) (hereinafter Chao et al) discloses the effects on nucleation rate and crystal growth, of varying the initial SiO.sub.2 /Al.sub.2 O.sub.3 ratio, alkalinity, and reaction temperature during zeolite synthesis from Na.sup.+ /TPA.sup.+ aluminosilicate gels. While sulfuric acid is disclosed as one of the reagents used in the experimental section, neither the amount nor the manner in which it is used is reported. Chao et al propose that alkalinity of the hydrogel affects the nucleation rate through two mechanisms, namely (1) the dissolution of the gel materials and formation of Al(OH)n, and (2) the polymerization of dissolved silicate and aluminate ions to form aluminosilicate or polysilicate ions which can act as a source of nuclei. From the data presented, the authors propose that increasing the alkalinity of the reaction mixture (a) increases dissolution of silicate species of the hydrogel, thereby shortening the induction period (i.e increasing nucleation rate), but (b) eventually results in restriction of the aforedescribed polymerization thereby lengthening the induction period at very high alkalinity. Chao et al therefore conclude that to achieve the highest nucleation rate an optimum alkalinity can be established where the dissolution and polymerization phenomenon are maximized. On the other hand, alkalinity is said to have almost no effect on the rate of crystal growth. The SiO.sub.2 /Al.sub.2 O.sub.3 ratio is alleged to have a two fold effect on reaction kenetics, namely, (1) except at low alkalinity, the lower the ratio (i.e. more aluminum) the higher the alkalinity needed to attain the aforedescribed optimum alkalinity point (since aluminum consumes OH.sup.- ions forming Al(OH)n) and (2) at low levels of alkalinity, the higher the ratio the faster the crystal growth rate. For an aluminum and sodium free system, excess TPAOH was required to achieve the comparable levels of alkalinity to compensate for omission of sodium hydroxide. The sodium/aluminum free system, however yielded only 16% crystallinity (see Table 3). While alkalinity of the reaction system is discussed in great detail, the initial pH of the reaction mixtures associated with the various alkalinities disclosed is never mentioned. Furthermore, the catalyst performance of the zeolites prepared by Chao et al was never tested and hence there is no correlation between alkalinity and/or initial pH on catalyst performance.
The crystallization kinetics of the NH.sub.4.sup.+ /TPA.sup.+ system were further studied in the paper "Synthesis and Growth of Zeolite (NH.sub.4, TPA)-ZSM-5 " by N. Ghamami, and L. Sand, Zeolites Vol. 3, pp. 155-62 (April 1983) (hereinafter Ghamami et al). The use of ammonium hydroxide is implemented instead of an alkali metal cation to eliminate the need for an ion-exchange step for subsequent conversion of ZSM-5 catalyst to the hydrogen form (e.g. typically Na.sub.+ is exchanged for NH.sub.4.sup.+ and the resulting material calcined to evolve NH.sub.3, to produce H-ZSM-5, and decompose the organic cation). In a system using precipitated silica powder, 25% TPAOH and initial SiO.sub.2 /Al.sub.2 O.sub.3 =28, the reaction does not proceed or proceeds slowly. Increasing the initial SiO.sub.2 /Al.sub.2 O.sub.3 ratio to 59 gives successful crystallization. This ratio is then used to explore the effect of varying the NH.sub.4.sup.+ /NH.sub.4.sup.+ +TPA.sup.+ ratio on crystallization. As this latter ratio is decreased (i.e. by increasing TPA.sup.+ and reducing NH.sub.4.sup.+ correspondingly) the nucleation and crystallization rates are found to increase. A decrease in the NH.sub.4.sup.+ /NH.sub.4.sup.+ +TPA.sup.+ ratio also corresponds to an increase in alkalinity which accelerates the reactant dissolution processes. Omitting NH.sub.4.sup.+ altogether (i.e., using TPAOH alone) results in spherical crystal aggregates while omitting TPA.sup.+ (i.e. using NH.sub.4 OH alone) results in an amorphous material (Compositions VI and VII respectively). At 180.degree. C. reaction temperature and a TPA.sup.+ /NH.sub.4.sup.+ ratio of 5/5, increasing the SiO.sub.2 /Al.sub.2 O.sub.3 ratio of the reaction mixture in the regime of 59; 69; 90 and alumina free, increases the nucleation and crystallization rates. The pH of the reaction mixtures employed in the first part of this paper (i.e. Compositions I to IX) is never reported. In the second part of the paper, TPABr is employed at the TPA.sup.+ source, Ludox AS40 (aqueous colloidal silica) as the silica source, and Reheis F-2000 aluminum hydroxide gel powder as the alumina source. The use of an initial SiO.sub.2 /Al.sub.2 O.sub.3 ratio of 59, and TPABr, rather than TPAOH, reduces the alkalinity of the reaction mixture producing an amorphous material at NH.sub.4 OH/TPABr ratios of 1.5 to 10 (Composition II). The use of excess NH.sub.4 OH (i.e. NH.sub.4 OH/TPABr=15; Composition XIV) gives a reaction mixture pH of 12-12.5 and produces euhedral crystals after 5 days. When the initial NH.sub.4 OH/TPABr ratio is reduced from 15 (in Composition XIV) to 12.5 (Composition XVI) thereby presumably reducing the initial pH to slightly below the 12-12.5 pH value (of Composition XIV), only 50% of the product is crystalline after 5 days.
It is appropriate to mention that the only initial reaction mixture pH reported in Ghamami et al is that of Composition XIV prepared in the absence of sodium. This pH value is not actively controlled (e.g. with acid), but is merely a result of the conditions established from the initial amounts and identity of ingredients selected. The TPABr salt is essentially neutral in terms of its effect on pH, and when a mixture containing a TPABr/NH.sub.4 OH/H.sub.2 O mole ratio of 8:120:750 was prepared, the pH of this mixture was 14. However, Ludox AS40, which has a pH of 9.2, can influence the initial pH of the reaction mixture.
Likewise Reheis alumina exhibits a pH of 8.6 and its addition to the reaction mixture can also affect the initial pH of the same. Consequently, it has been concluded that the identity of the source of the alumina and silica in Composition XIV of Ghamami et al is responsible for the inherent initial 12.5 pH of the same.
The article, "Preparation of Zeolite Catalyst for Synthesis of Lower Olefins from Methanol" by E. Kikuchi, R. Hamana, S. Hamanaka and Y. Morita, J. of Japanese Petroleum Institute, Vol. 24, pp. 275-280 (1981) (hereinafter Kikuchi et al) discloses the preparation, and testing for methanol conversion, of ZSM-5 catalysts. Kikuchi et al examine two catalysts designated A and B. Catalyst A is prepared in accordance with the standard Mobil ZSM-5 technique of U.S. Pat. No. 3,702,886, using silica gel, TPA.sup.+ and NaAlO.sub.2. Catalyst B is prepared using water glass (92.9% SiO.sub.2, 9% Na.sub.2 O), aluminum nitrate and TPA.sup.+. However, sufficient 1N HNO.sub.3 is added to the reaction mixture for Catalyst B to bring the reaction mixture pH to 10-10.5. As the pH is reduced, a gellous solution forms and is stirred. Catalyst samples A and B are then tested for methanol conversion with further testing of Catalyst B at varying SiO.sub.2 /Al.sub.2 O.sub.3 ratios. Comparing Catalysts A and B on a morphological basis, Kikuchi et al report that the size of the crystallites of Catalyst B is about 4 times that of Catalyst A and that the crystallinity of Catalyst B after 1 day of crystallization is about the same as Catalyst A after 6 days. In terms of catalyst performance, Catalyst B is said to show a selectivity to lower olefins about 1.5 times that of Catalyst A at similar conversion levels. Kikuchi et al conclude that the differences in activity may be attributable to a slight difference in pore structure which cannot be identified by XRD, which in turn may enhance the rate of diffusion of the olefin out of the pores. Increasing the SiO.sub.2 /Al.sub.2 O.sub.3 ratio of Catalyst B in the regime of 50; 202; 362; and 602 results in an increase in selectivity to lower olefins, a decrease in the activity of catalyst, and a decrease in the selectivity to aromatic hydrocarbons. Note that Kikuchi et al do not specify whether the SiO.sub.2 /Al.sub.2 O.sub.3 ratios reported are those of the actual zeolite, or the starting ratios employed in the reaction mixture. It is further noted that while Kikuchi et al appear to be the first workers to employ active pH control with an acid, in a sodium cation containing system, the pH to which the reaction mixture was adjusted is confined to 10-10.5 with no recognition of any influence of such control on catalyst life.
European Patent Application 93,519 discloses a process for preparing high silica containing zeolites of the ZSM-5 family wherein a buffer is employed to control the pH of the reaction mixture during crystallization between 9.5 and 12. This process is said to be based on the discovery that the final pH of the reaction mixture, will determine the morphology of the resulting crystals. More specifically, a final pH of 10-10.5 is said to produce rod-shaped crystals, a final pH of 12 to 12.5 twinned short prismatic crystals with near spherulitic morphology, and a final pH of 11 to 12, a morphology intermediate between the above noted morphologies. The reaction mixture which is associated with the above morphologies contains water, a source of quaternary ammonium cations, silica, and an alkali metal. An aluminum source is optional. No utility is disclosed for the zeolites prepared in accordance with this process and consequently the activity of such zeolites was never tested for any purpose. The buffers disclosed at page 3 are conjugate bases, i.e. salts, of a weak acid and a strong base. In contrast, the present invention excludes the presence of buffers from the reaction mixture. Furthermore, the activity of the catalysts of the present invention has been found to be dependent on the initial pH of the reaction mixture and hence there is no need to employ a buffer. Furthermore, there is no recognition of the effect of adjusting the initial pH of the reaction mixture on catalyst life in accordance with the process of the present invention.
U.S. Pat. No. 4,275,047 discloses a process for preparing zeolites such as ZSM-5 wherein the use of alkylammonium ions can be avoided by inclusion in the reaction mixture of a seed zeolite having a specifically defined pore diameter. For unspecified reasons, the reaction mixture is disclosed as preferably containing one or more anions of strong acids, especially chloride, bromide, iodide or sulphate. Such anions can be introduced as an acid, and/or alkali metal, aluminum, ammonium or onium salts. Note also that H.sub.2 SO.sub.4 is employed in Example 5, it is assumed, as a source of sulphate ions. No initial reaction mixture pH is specified.
In the article "Synthesis and Characterization of ZSM-5 Type Zeolites III, A Critical Evaluation of the Role of Alkali and Ammonium Cations" by Z. Gabelica, N. Blum, and E. Derouane, Applied Catalysis, Vol. 5, pp. 227-48 (1983) (hereinafter Gabelica et al), the role of alkali metal and ammonium cations in the nucleation and growth of ZSM-5 zeolites is studied. The authors conclude that the morphology, size, chemical composition and homogeneity of the crystallites depend on competitive interactions between TPA.sup.+ or alkali metal cations and aluminosilicate polymeric anions during early stages of nucleation. The crystallization time of a Na free, TPA.sup.+ /NH.sub.4.sup.+ based system is reported as 93 days. While both acidic and basic systems are studied, the TPA is added as the bromide salt, and in some instances the pH of the reaction mixture is increased from acidic (e.g. 2-4) to basic (pH 9) by the addition of sodium silicate. One significant observation of Gabelica et al is that the initial SiO.sub.2 /Al.sub.2 O.sub.3 ratio appeared to have little influence on the final zeolite composition, the later being strongly dependent on the nature of the alkali counterion, which in turn affected the size of the crystallites. None of the Gabelica et al zeolites are tested for catalyst performance.
Commonly assigned U.S. patent application Ser. No. 630,723, filed July 13, 1984 by A. Bortinger, W. Pieters, and E. Suciu, is directed to a process for preparing zeolites, e.g. ZSM-5, on an active initial pH controlled basis (i.e. initial pH=9.0 to 12.5) but in the substantial absence of alkali metal cations (e.g. sodium) in the reaction mixture.
From the above discussion it can be seen that except for the above discussed commonly assigned Bortinger et al application, ZSM-5 zeolites have been synthesized with organic cations under controlled processing conditions on a sodium free basis, but in the absence of active initial pH control, although the catalyst performance of such zeolites does not appear to have been tested. On the other hand the use of active initial pH control with an acid has only been applied to sodium-TPA.sup.+ binary cation conventional ZSM-5 systems at a pH of 10-10.5 and the resulting zeolite catalyst exhibits good catalyst performance relative to the absence of pH control.
However, to the best of the inventors' knowledge herein, the combination of the use of sodium containing cation reaction system under strict active initial pH controlled conditions (i.e., pH 11.3 to 11.7) bin accordance with the process of the present invention has never been reported, nor has the catalyst performance of zeolites prepared in this manner.